Sponsored Links

Speedtests - upload protocol

I have just run a speedtest, using the well-known speedtest.net. It shows 17.41 down and 0.86 up, which is fairly typical. The download speed roughly matches what I get from decent sites if I actually time a biggish download, but...

The upload speed (which, on umpteen tests on speedtest.net and other test sites never goes above 0.87) does not match what I see when I shove my roughly 25MB patch file for the database onto my website about twice a week. My FTP programme has a little timer thingy, which consistently shows in excess of 1000 - usually around 1015 - and actual time confirms that there's nothing wrong with it if you sit down with a calculator and work it out. Only once in the last six months has it ever shown less than 1000, and that wasn't much below, can't remember the exact figure, but something in the 990s.

My theory is that this is because all the online speedtests do the same as for the download - they measure http not ftp.
Does that sound a reasonable idea to our more technologically clued-up members?
 
In "excess of 1000" what? Kbps, KBps.. other?

Leaving that aside, it is worth remembering that FTP is a variable environment and a Speedtest is usually a fairly linear one that most likely only uses compressed information, which is often transferred in binary format, for an "accurate" reading.

On the other hand FTP will transfer in different ways, such as ASCII or Binary, and often the data you send will also be in an uncompressed format that can have a big effect on what you see depending upon how the server is setup. The FTP software will usually adjust automatically to the file format and database files (assuming SQL here) are usually ASCII not Binary.

It is also true that service/protocol type (FTP, HTTP etc.) can have an effect but that does rather depend upon how your ISP, software and routers network management setup treats the different protocols. Anyway, my point is that trying to pin down precisely what makes the small difference is very difficult but you're not really comparing Apples with Apples. I find it best not to dwell on things like this too much :) .
 
For what it is worth, speedtest.net does an http POST for it's upload test

Easy to find out using Ethereal/ wireshark http://www.wireshark.org/


POST /speedtest/upload.php?x=0.5393625902943313 HTTP/1.1
User-Agent: Opera/9.80 (Windows NT 6.1; WOW64; U; IBM EVV/3.0/EAK01AG9/LE; en-GB) Presto/2.10.229 Version/11.64
Host: speedtest.krystal.co.uk
Accept: application/xhtml+voice+xml;version=1.2, application/x-xhtml+voice+xml;version=1.2, text/html, application/xml;q=0.9, application/xhtml+xml, image/png, image/webp, image/jpeg, image/gif, image/x-xbitmap, */*;q=0.1
Accept-Language: en-GB,en;q=0.9
Accept-Encoding: gzip, deflate
Referer: http://c.speedtest.net/flash/speedtest.swf?v=302210
Connection: Keep-Alive
Content-Length: 145181
Content-type: application/x-www-form-urlencoded

content0=TQJNDYHOKPYAIXQKDIVABFPHZAGWVYCCTDYQSWBYPIJOVKJZESQAIJDZLGWVRCIAAFEXYASRSWMQSSMEWGJQLGLCNCMCJOLMMZNPNONHEBY...
 
Sponsored Links
@ Mark...

In "excess of 1000" what? Kbps, KBps.. other?
Sorry - Kbps.
At some point (presumably at 1024), the display changes to Mbps, so on a really good day it'll say 1.03 etc. - very occasionally it gets to 1.05 - but usually only for comparatively small files, so I don't count that.

database files (assuming SQL here) are usually ASCII not Binary.
It's not SQL - it's a bit of an oddity, a genealogy programme (GeneWeb) that runs as a "service" on my own PC (I have to kick "gwd", the geneweb "daemon" into action to use it) but uses a cgi script on the website. I'm not sure, but I think the answer is binary.

I find it best not to dwell on things like this too much :).
I'm not letting it keep me awake at nights! :laugh:
It just struck me as odd that there was always this difference. About 1.00 compared to 0.86 is not what I'd call a "small" difference personally - and being the inquisitive type, I was curious as to why, so did a little head scratching and thought I might have found the reason. Obviously, from your other comments, things are more complex than I realised.

@ Mel...

Thanks, Mel - at least that confirms that they and I are using measurements based on different protocols, so I now know that I was right about that bit, even if it isn't the actual reason for the consistent difference. :)
 
Different protocols have different overheads (error checking etc), have you considered that.
 
No, I hadn't - the difference between them is not something that I pretend to understand at all, I was just curious as to whether the use of the different protocols (whatever they do!) might be the reason for the different speeds.
 
Top
Cheap BIG ISPs for 100Mbps+
Community Fibre UK ISP Logo
150Mbps
Gift: None
Virgin Media UK ISP Logo
Virgin Media £22.99
132Mbps
Gift: None
Vodafone UK ISP Logo
Vodafone £24.00 - 26.00
150Mbps
Gift: None
NOW UK ISP Logo
NOW £24.00
100Mbps
Gift: None
Plusnet UK ISP Logo
Plusnet £25.99
145Mbps
Gift: £50 Reward Card
Large Availability | View All
Cheapest ISPs for 100Mbps+
Gigaclear UK ISP Logo
Gigaclear £17.00
200Mbps
Gift: None
Community Fibre UK ISP Logo
150Mbps
Gift: None
Virgin Media UK ISP Logo
Virgin Media £22.99
132Mbps
Gift: None
Hey! Broadband UK ISP Logo
150Mbps
Gift: None
Youfibre UK ISP Logo
Youfibre £23.99
150Mbps
Gift: None
Large Availability | View All
Sponsored Links
The Top 15 Category Tags
  1. FTTP (6026)
  2. BT (3639)
  3. Politics (2721)
  4. Business (2439)
  5. Openreach (2405)
  6. Building Digital UK (2330)
  7. Mobile Broadband (2146)
  8. FTTC (2083)
  9. Statistics (1901)
  10. 4G (1816)
  11. Virgin Media (1764)
  12. Ofcom Regulation (1582)
  13. Fibre Optic (1467)
  14. Wireless Internet (1462)
  15. 5G (1407)
Sponsored

Copyright © 1999 to Present - ISPreview.co.uk - All Rights Reserved - Terms  ,  Privacy and Cookie Policy  ,  Links  ,  Website Rules